Decrease write quorum for fully-replicated tables (buckets, API keys, ...) #919
No reviewers
Labels
No labels
action
check-aws
action
discussion-needed
action
for-external-contributors
action
for-newcomers
action
more-info-needed
action
need-funding
action
triage-required
kind
correctness
kind
ideas
kind
improvement
kind
performance
kind
testing
kind
usability
kind
wrong-behavior
prio
critical
prio
low
scope
admin-api
scope
background-healing
scope
build
scope
documentation
scope
k8s
scope
layout
scope
metadata
scope
ops
scope
rpc
scope
s3-api
scope
security
scope
telemetry
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference: Deuxfleurs/garage#919
Loading…
Reference in a new issue
No description provided.
Delete branch "hotfix/1.0.0-rc1-red-ftr-wquorum"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Currently, the write quorum for fully-replicated tables (buckets, access keys, etc) is n-1 nodes, so we tolerate a single node failure. Unfortunately this means that if a whole zone is down that contains several nodes, admin operations (creating buckets, etc) stop working. This is not normal as we expect Garage to continue operating normally when a whole dc is down.
This fixes the issue by reducing the write quorum to a majority of nodes. This patch is used in production at Deuxfleurs.
TODO: what are the consistency implications of this change?
View command line instructions
Checkout
From your project repository, check out a new branch and test the changes.Merge
Merge the changes and update on Forgejo.